Biochem. J. (1988) 251, 73-79 (Printed in Great Britain)

Complete amino acid sequence of p453-plasmid-mediated PIT-2fI-lactamase (SHV-1)

Michel BARTHELEMY, Jean PEDUZZI and Roger LABIAMuseum National d'Histoire Naturelle, C.N.R.S. U.A. 401, 63 Rue Buffon, 75231 Paris Cedex 05, France

The complete amino acid sequence of the p453-plasmid-mediated PIT-2 ,-lactamase (SHV- 1) wasdetermined. The protein contains 265 residues. Peptides resulting from digestions with trypsin,Staphylococcus aureus V8 proteinase, chymotrypsin and Lys-C proteinase and cleavage with CNBr wereseparated and purified by using reverse-phase h.p.l.c. The amino acid sequence of each peptide was manuallydetermined with the dimethylaminoazobenzene isothiocyanate/phenyl isothiocyanate double-couplingmethod. The primary structure of PIT-2 fl-lactamase was compared with those of two closely relatedenzymes, namely TEM-1 8-lactamase and the 8-lactamase of Klebsiella pneumoniae strain LEN-1. ThePIT-2 ,?-lactamase amino acid sequence was strongly retained, with respectively 68 % and 88 % homology.Thus PIT-2 enzyme could represent an evolutionary step between a chromosomally encoded ,1-lactamaseand the plasmid-mediated TEM ,1-lactamases.

INTRODUCTIONThe /-lactamases, which include a large number of

chromosomally and plasmid-mediated enzymes, play amajor role in bacterial resistance to penicillins andcephalosporins. On the basis of their biochemicalproperties and primary structures, /8-lactamases havebeen divided into three evolutionary distinct classes(Ambler, 1980; Jaurin & Grundstrom, 1981; Bergstromet al., 1982). TEM-type enzymes, the most widespread,-lactamases (Matthew, 1979), belong to class A andare plasmid-encoded. Plasmid-determined ,-lactamaseshold an important place in bacterial resistance since theyare widely distributed throughout the world in bacterialspecies (Matthew, 1979; Medeiros, 1984). More than 20such well-characterized enzymes have been reported(Medeiros et al., 1985). Among them, a penicillinaseencoded by a transposon (Nugent & Hedges, 1979),frequently encountered in Klebsiella pneumoniae where itwas first detected by Pitton (1972), was named PIT-2 byO'Callaghan et al. (1978) and further designated SHV-Iby Matthew et al. (1979). On the basis of determinationofkinetic constants and amino acid composition analysis,PIT-2 enzyme appeared to be closely related to TEM-type penicillinases (Barthelemy et al., 1986), although noevolutionary relationship was found between them byDNA hybridization analysis (Levesque et al., 1987). Animportant feature of PIT-2 fl-lactamase was described byKliebe et al. (1985), who isolated laboratory mutantsthat showed high broad-spectrum hydrolytic activitytowards cephalosporins. Moreover, a clinical isolate ofcefotaxime-resistant Klebsiella ozaenae was also found toproduce a transferable ,-lactamase (called SHV-2) thatcould be a natural mutant of PIT-2.

This set of properties of PIT-2 fl-lactamase made it anattractive candidate for structural analysis. We havepreviously published the determination of the first 107residues of the N-terminal sequence (Barthelemy et al.,1987). In the present paper the complete amino acidsequence ofPIT-2,f-lactamase is described and comparedwith those of other related enzymes.

EXPERIMENTALPurification of the PIT-2 fp-lactamaseThe PIT-2 fl-lactamase was purified from a

hyperproducing variant of an Escherichia coli K12strain carrying plasmid p453, as previously described(Barthelemy et al., 1986). The protein was found tobe homogeneous by SDS/polyacrylamide-gel electro-phoresis and analytical isoelectric focusing.

CarboxymethylationThe ,-lactamase, dissolved in 6 M-guanidinium

chloride/2 mM-EDTA/0.1 M-Tris/HCl buffer, pH 8.3,was reduced with a 3-fold molar excess of dithiothreitolfor 1 h at room temperature under N2. A 6-fold molarexcess of iodoacetic acid over dithiothreitol was addedand the solution was left under N2 for 1 h in the dark atroom temperature. After the addition of I % (v/v) of 2-mercaptoethanol the S-carboxymethylated ,J-lactamasewas extensively dialysed against 50 mM-NH4HCO3/0.01.% (v/v) thiodiglycol and then freeze-dried.

H.p.l.c. purification of proteinase-digest peptides andCNBr-cleavage fragments

All the peptides were purified by h.p.l.c. with an LKB2152 controller, two LKB 2150 pumps, an LKB 2238detector and a Rheodyne injector. Peptides separated ona Waters C18 ,uBondapak column (10,m particle size,4 mm x 300 mm) were detected at 226 nm and collectedmanually. The flow rate was 1 ml/min at room tempera-ture. Peptide elutions were performed with 0.1 % (v/v)trifluoroacetic acid in water (solvent A) and acetonitrile(solvent B). A three-step linear gradient from solvent Ato solvent B was used. In some instances peptides werefurther chromatographed with a linear gradient from0.25 mM-ammonium acetate buffer, pH 6.0, to 0.5 mm-ammonium acetate buffer (pH 6.0)/acetonitrile (2:3,v/v). Finally all peptides were rechromatographed withisocratic elution with 0.1 % (v/v) trifluoroacetic acid inacetonitrile.

Vol. 251

73

M. Barthelemy, J.- Peduzzi and R. Labia

Tryptic and chymotryptic cleavagesThe digestion of the S-carboxymethylated protein

(90 nmol) by trypsin (Sigma Chemical Co.) and of thelong staphylococcal-proteinase-digest peptide V13(10 nmol) by chymotrypsin (Sigma Chemical Co.) wereperformed for 4 h and 2 h respectively at 37 °C in 0.25 M-NH4HCO3 at an enzyme/substrate ratio of 1: 50(w/w).

Staphylococcal-proteinase digestionsThe digestion of the S-carboxymethylated /-lactamase

(90 nmol) and tryptic peptide Ti 1 (25 nmol) by Staphylo-coccus aureus V8 proteinase (Sigma Chemical Co.) wereperformed for 8 h at 37 °C in 0.25 M-NH4HCO3 at anenzyme/substrate ratio of 1: 25 (w/w). A second equiva-lent portion of proteinase was added to the peptide T 11solution, and the reactions were allowed to proceed foran additional 12 h.

Lysine-specific proteolytic cleavageThe long staphylococcal-proteinase-digest peptide

VI 5 (20 nmol) was dissolved in 8 M-urea and thendigested in 4 M-urea/0.25 M-ammonium bicarbonatebuffer, pH 9, by endoproteinase Lys-C from Lysobacterenzymogenes (Boehringer Mannheim) at 37 °C for 18 hat an enzyme/substrate ratio of 1:100 (w/w).

CNBr cleavageS-Carboxymethylated protein (75 nmol) dissolved in

500 #1 of 70 % (v/v) formic acid was digested with CNBr(7.5 mg) for 20 h at 37 'C. In order to convert C-terminalresidues into a single form (homoserine lactone) thepeptide mixture was treated before h.p.l.c. analysis withanhydrous trifluoroacetic acid for 1 h at 20 'C (Ambler& Brown, 1967).

Carboxypeptidase digestions of /)-lactamase and trypticpeptide T26

S-Carboxymethylated ,-lactamase (10 nmol) ortryptic peptide T26 (15 nmol) in 1 ml of 0.2 M-N-ethyl-morpholine/acetate buffer, pH 8.5, containing 20 nmolof norleucine as an internal standard, was digested at25 °C with a mixture of carboxypeptidases A and B(Boehringer Mannheim). At specified times (30 min to4 h) samples (200 1u) were withdrawn and subjected toamino acid analysis.

coupling method, essentially as described by Allen (1981),with slight modifications. After the coupling step theexcess reagents were removed by three succesive extrac-tions with 600 ,u of heptane/ethyl acetate (3: 1, v/v). Thecleavage step was performed with anhydrous trifluoro-acetic acid at 58 °C under N2 for 5 min and for a further5 min if proline was the expected N-terminal residue.

Peptide nomenclaturePeptides were numbered in the order in which they

occur in the sequence starting from the N-terminus.Tryptic peptides are designated by a T, S. aureus V8proteinase peptides with a V, Lys-C proteinase peptideswith a K, chymotryptic peptides with a C and CNBr-cleavage peptides with CN. Peptides obtained fromsubdigestions are designed by indicating first the parentpeptide followed by a second letter indicating the cleavagemethod and numbered according to their position in thesequence of the parent peptide.

RESULTS

The N-terminal analysis ofthe purified protein revealedonly one amino acid derivative at each position amongthe first six residues. Following analyses were perturbedby non-specific cleavages of peptide bonds, whichoccurred during the trifluoroacetic acid treatment. Alimitation of the reaction time for the cleavage stepallowed identification of the first 15 N-terminal aminoacid residues (Barthelemy et al., 1987). The wholesequence of PIT-2 ,J-lactamase was established by

0.4

0.2

Cm-,

6L)4-

0

C.)C0u

Amino acid analysisPeptides (1-2 nmol) were hydrolysed at 110 °C under

vacuum in the presence of 200 ,1 of 6 M-HCI for 23 h.For large peptides amino acid analysis was performedfor 48 h and 72 h. Hydrolysates of peptides from CNBrcleavage were treated-with aq. 10% (v/v) pyridine/aceticacid buffer, pH 6.5, at 105 °C for 1 h, to allow conversionof homoserine lactone into homoserine (Ambler, 1965).Amino acids were separated by using a Kontron Liquimat2 amino acid analyser incorporating a Waters 420fluorimeter and detected after the formation of deriva-tives with o-phthalaldehyde.

Amino acid sequence determinationsSequential degradation of 2-10 nmol of peptides was

performed according to Chang's manual diaminoazo-benzene isothiocyanate/phenyl isothiocyanate double-

0

Time (min)Fig. 1. Separation of tryptic-digest peptides from PIT-2 p-

lactamase by h.p.l.c.The tryptic digest of PIT-2 enzyme (90 nmol) adjusted topH 2 with trifluoroacetic acid was filtered through a0.45 /,m-pore-size filter, then concentrated to 400 ltl.Portions (100 1dl) were applied to a C18,uBondapak columnand eluted as described in the text. , A226; * , concn.of solvent B. The numbering of the peaks refers tothe numbering of the corresponding peptides describedunder 'Peptide nomenclature' in the Experimental section.Those peaks that ass indicated by two numbers wererechromatographed at pH 6.

1988

74

la

Amino acid sequence of PIT-2 ,3-lactamase (SHV-1)

-l-

I., -_ I.,

I--, z _ _

_ _

1

_

-

__-- -L.2 -_9O F9Cl-. .

-7

0 0

I-- I--

o) o

'Ir dR " - -m " -r,- I.,.,,- .~I.- .~.- .-

.

I--

l- _- t oo O-C Cl--

1-1

o Cl

_ I-, _:7- o> _ ;

Cl _ ^ - 00 -

O~~ ON 00 - N

L.7 -0----L

CL.1N -- - 0 -LWo

A _Cl _ N Cl_6 - 7 . .._.I .- , .- I0O eCNc~r-:r-:C~ Clc

11-11

1-1 1-11 1-1 I.-I

0OCNOC--; -; 6 4

-

O

Cl C - -

It ON ON- -- -0

-z_ - -~~~~~~~~~~= -7-

- --L1 --C) cr o; C o

- -- -_ _ _ Cl -

-7I -ZC

Cl*

_.- _.IC)

_- I-, _

C>ClON-:orCZ 65- _

I--,

x ,- v_ x at (,< U > . .1 a. Q-<

u

Ou0.)0)0

10

1._I-

._0

00.

0.0)

10

'IC

ICcliHONICl,.

Cl,en Cl

"

ClN

eCnCl Cl

-ON

Cl4

ON0

- ClClj

0

ON -l-

EH-0-

-,; _

H-

N --

H00

"-

N

-

HN

N

P_

"C/

_

0)0

cn

v _D

* c,

O.C4

M c o0 -_m , o'IC _r-_C .- M - l M C ON-a eM t m t_

0 N £c e " Cl0OoONo 00 000o o o o m N

"-- M - " - - n Cl4

- -L_N C-er_ "

z oo

1-_ __1- I-

_ _z

I--

I.-Cl-

C-N_1 I---

N9

_ _

N

Cl- -

1-1

1-

006

7- ~z unC-o

f

N N _-_ _-_ - -r> io -~ oo *

F _ _ _ o-~C

1-

Cl4ON

Cl -I-- _~

n o. C.7~C' -f90

I-

o.

~-C- 00 -

-. i

_ O Cl

O 0006 i

0

-0 Cl

--

_- _ ,1

-O.Z

_- _-

11_ I--,__

I--,- Cl I-_- I.- ._ _~ .-

_ __ _ _ ,)9 ON cO 0 ONON ON 0

z-~: 0 ~--~ -~=~ 00 0~ -~C

_ __ - -~~-*CD

ON ONON 00 0 N6 66 6

oo-

ON 000 -- -;

,IC

0 Zec

o-z ONONo o

__ __

ON 00_-- -

ONON9Z00-

C'N

-ON..- . 00-e.N- Cl

-.- ...0rCN

.r0N. ZOCl

Col - _-

00 00 Cl

0 - --1

oo O *>-

_4 - _;O~

N 00 _

o.I Zr- C.O _ -z

0~~~~~~~~~~~V. u -O

)-l<; d. 1 4 0

~ = 0u ,C CAE~r) ( -.I-

75

000)

00S-0)

-0

0._

.-_

CZ

0

0

0.

0

0._

en

M-

-

- Cl

00

_N

ON N

r00

o 00

r-

CfN

'r-

HfN

00

r IC

I^F- (7_

00

0

-ON

H-

-a-

_.,

C0)

CZ-

-' o

o S

. oz

o Q

- S

.s- .cj

ct

ct

*--~ Q

oE

Q *o

0.-

CZ-o Ct

L O

o. 0C

0- 0o -0)_ ) .

C/C

0)

I-

._'

E0

G,

0)

E.c50x

0)

C/

2CucnCU

Ei

85

0

C,'

0

0r

Vol. 251

11-1

=1 1-11 11-11 11-11- 1-11- - - _,, en r1r)1-1 1-1 1-1 1-1 1-1

- C. CD - C. CD ---; 6 --; --; .. C,. C,.

76 M. Barthelemy, J. Peduzzi and R. Labia

10 20 30

Ser-Pro-Gln-Pro-Leu-Glu-Gln-Ile-Lys-Leu-Ser-Glu-Ser-Gln_Leu-Ser_Gly_Arg-Val-Gly-Met-Ile-Glu-Met-Asp-Leu-Ala-Ser-Gly-Arg-

Whole protein

Tl L- T2 T3

vi V2 V3 V4

40 50 60

Thr-Leu-Thr-Ala-Trp-Arg-Ala-Asp-Glu-Arg-Phe-Pro-Met-Met-Ser-Thr-Phe-Lys-Val-Val-Leu-Cya-Gly-Ala-Val-Leu-Ala-Arg-Val-Asp-

T4 T5 T6 T7-V4 V7

70 80 90

Ala-Gly-Asp-Glu-Gln-Leu-Glu-Arg-Lyi-Ile-flis-Tyr-Arg,-Gln-Gln-Asp-Leu-Val-Asp-Tyr-Ser-Pro-Val-Ser-Glu-Lys-His-Leu-Ala-Aisp-

T7 T8 T10 Tll

V7- J V8 V9 V10

100 110 120

Gly-Met-Thr-Val-Gly-Glu-Leu-Cys-Ala-Ala-Ala- Ile-Thr-Met-Ser-Asp-Asn-Ser-Ala-Ala-Asn-Leu-Leu-Leu-Thr-Ala-Val-Gly-Gly-Pro-

Tll _----_-_ ------_ ---__-_-__-_-__

Tll-V4T11-V5 -

V1o V13

130 140 150

Ala-Gly-Leu-Thr-AIa-Phe-Leu-Arg-Gln-Ile-Gly-Asp-Asn-Val-Thr-Arg-Leu-Asp-Arg-Trp-Glu-Thr-Glu-Leu-Asn-Glu-Ala-Leu-Pro-Gly-

--------------- Tll -----------I T12 T13

-T1l-V4--------------------JTT- -V5 J

V13 ---J-- V14 V15

V13-C J

160 170 180

Asp-Ala-Arg-Asp-Thr-Thr-Thr-Pro-Ala-Ser-Met-Ala-Ala-Thr-Leu-Arg-Lys-Leu-Leu-Thr-Ser-Gln-Arg-Leu-Ser-Ala-Arg-Ser-Gln-Arg-

T13 T14 T16

V15 -

I V15-K2

190 200 210

Gln-Leu-Leu-Gln-Trp-Met-Val-Asp-Asp-Arg-Val-Ala-Gly-Pro-Leu-1le-Arg-Ser-Val-Leu-Pro-Ala-Gly-Trp-Phe-I le-Ala-Asp-Lys-Thr-

T18 T19 T20 -

.-- - - - - - - - - - - - - - - - - - - - - - - - - - - - V 5 - - - - - - - - - - - - - - - - - - - - - - - -----V15

--------------------------- V15-K2 ----------------------------------- -CN4

220 230 240

Gly-Ala-Gly-Glu-Arg-Gly-Ala-Arg-Gly-Ile-Val-Ala-Leu-Leu-Gly-Pro-Asni-Asn-Lys-Ala-Glu-Arg-I le-Val-Val- I le-Tyr-Leu-Arg-Asp-

T20 T22 T24- ------------------------------------------------ _ _-_-----------_-_-_-_-_-_-_-_-_-----Vi1-

V15-K3 V15-K4

----------------------------------------------------------- CN4 --------------------------------------------------------

250 260 265

Thr-Pro-Ala-Ser-Met-Ala-Glu-Arg-Asn-Gln-Gln-Ile-Ala-Gly-I le-Gly-Ala-Ala-Leu-I le-Glu-iIis-Trp-Gln-Arg

T25 T26----- V15 ----------- V16 V1i7

-V15-K4 -

-------- CN4 ------ I CN5

1988

Amino acid sequence of PIT-2 fl-lactamase (SHV-I)

analysis of peptides generated by a tryptic digestion. Thealignment of tryptic peptides was found by usingoverlapping fragments isolated from staphylococcal-proteinase and CNBr cleavages or from subfractionationof large peptides by chymotrypsin, Lys-C proteinase andstaphylococcal proteinase. All peptide mixtures weresubmitted to reverse-phase h.p.l.c. A typical elutionpattern, corresponding to the tryptic digest, is indicatedin Fig. 1. The amino acid compositions of the 26 trypticpeptides are listed in Table 1. Except for peptide Ti 1, allthe peptides were completely sequenced by Edmandegradation. However, the C-terminal lysine residues ofpeptides TI, T5 and T10 were not identified. Theunsequenced part of peptide Ti 1, of which only the first28 residues were determined, was elucidated by analysisof subpeptides TiI-VI to TiI-V5 obtained fromfractionation of peptide T 11 with an excess of S. aureusproteinase. The results of sequences analysis are given inFig. 2. By comparison with the N-terminal sequence ofthe intact S-carboxymethylated protein, we positionedpeptide TI at the N-terminus. As all peptides ended in anarginine or lysine residue, because of the tryptic cleavage,no particular one could be suspected to be located at theC-terminus. Peptides T9, T15, T17, T21 and T23, whichresult from partial cleavages at lysine residues, are notshown, but their positions in the sequence are indicatedin Table 1. Three small peptides (residues 174-177,178-180 and 216-218), which most probably were elutedwith the buffer peak, were not recovered by h.p.l.c. TheArg-Phe bond at position 40-41 and the Arg-Trp bondat position 139-140 were apparently resistant to trypticcleavage.

Staphylococcal V8 proteinase cleavages mainly oc-curred at peptide glutamyl bonds, although one aspartylbond (position 76-77) was fully hydrolysed (Fig. 2).Peptide V6 resulted from an unusual cleavage of aThr-Ala bond in position 33-34. Another Thr-Ala bondcleavage (position 115-116) was also observed duringsubfractionation of tryptic peptide Ti 1 by staphylococcalproteinase. H.p.l.c. analysis of a reaction mixture sampleremoved after 1 h digestion revealed preferential peptideglutamyl bond cleavages at residues 143, 146, 247 and261. That, mainly, yielded the fragment V15, whichformed an insoluble precipitate, even in the presence of4 M-urea. So the aggregation of peptide VI5 did notallow the Glu-Arg bond proteolysis at positions 214-215and 231-232. Partial amino acid sequence of peptidesV13 and V15 (35 and 28 residues respectively) wereperformed, whereas all the other peptides were fullysequenced. Peptide V17 was suspected to be the C-terminal fragment of the ,J-lactamase as it ends in a non-specific arginine residue. Structural analysis of peptidesisolated from staphylococcal proteinase digestions pro-vided evidence for the alignment of tryptic peptidesspanning residues 1-131.

In order to elucidate the unsequenced part of peptideV13, the latter was subdigested by chymotrypsin, the

Fig. 2. Amino acid sequence of PIT-2 &I-actamaseResidues identified by manual sequence determinationsare underlined by continuous lines. Blanks denote residuesnot identified. Broken lines indicate peptides only checkedby amino acid composition analysis. C-Terminal residuesdetermined after carboxypeptidases digestion are un-derlined by-e.

reaction being stopped before complete digestion. Thepeptide V13-C was purified from the chymotrypticdigest by C18,uBondapak h.p.l.c. Its partial amino acidsequence (Fig. 2) overlaps tryptic peptides T12 and T13.Moreover the sequenced part of the N-terminal sequenceof peptide V15 spans peptides T13, T14, T15 and T16.Therefore the sequence of the first 174 N-terminalresidues of the protein was determined.From sequence analysis of tryptic peptides we knew

that the 101-residue peptide V15 contained three lysineresidues residing in strategic positions. In order to obtainoverlapping sequence at positions 170-248, the fragmentV15 was subjected to Lys-C endoproteinase digestion.Pure peptides VI5-KI to V15-K4 were removed from asingle reverse-phase h.p.l.c. Amino acid sequence analysisallowed identification of the first 24 residues of peptideVI 5-K2 (Fig. 2). The other three peptides were fullysequenced, although C-terminal lysine derivatives werenot identified. Sequencing of peptide VI 5-K2 establishedthe sequence at positions 174-180 and confirmed thestructure of fragments 168-173 and 181-191. Lastly, thepositions 216-218 were deduced from the structure ofpeptide V15-K3.

Furthermore fragments overlapping tryptic peptidesT18, T19 and T20 were obtained from CNBr cleavage ofthe PIT-2,f-lactamase. Five fragments at positions 1-21,25-43, 162-186, 187-245 and 246-265, isolated byreverse-phase h.p.l.c., spanned the first 43 N-terminalresidues and the last 105 C-terminal residues. Noadditional fragment from the protein core was recoveredpure in useful amount. The N-terminal sequences ofpeptides CN4 and CN5, which provide new data forstructure elucidation, are shown in Fig. 2.

Sequence determinations of subpeptides from Lys-Cproteinase digestion and of peptides from CNBr cleavageprovided overlaps for tryptic peptides T18 to T26. Thusthe PIT-2 ,J-lactamase sequence was extended fromposition 174 to residue 265.The C-terminal sequence of PIT-2 ,-lactamase was

investigated by the use of carboxypeptidases A and B.After incubation with carboxypeptidase B, only arginine(0.8 mol/mol of protein) was found in the digest. With amixture of the carboxypeptidases A and B, the aminoacids released were arginine, glutamine and tryptophan.It is noteworthy that no trace of histidine was detected.A time-course study of amino acid release yielded the C-terminal sequence -Trp-Gln-Arg, which likewise wasdeduced from digestion of the tryptic peptide T26 withcarboxypeptidases A and B. The results confirmed thefindings from Edman degradation of the proteolyticpeptides.

DISCUSSIONThe complete 265-residue amino acid sequence of

PIT-2 /1-lactamase is summarized in Fig. 2. The Mrvalue calculated from the sequence is 28 835. This is closeto the value of 27 500 that was obtained by SDS/poly-acrylamide-gel electrophoresis (Barthelemy et al., 1986).The amino acid composition determined from the puri-fied protein is in agreement with the sequence (Table 1).

In Fig. 3 the amino acid sequence of PIT-2 ,-lactamase is compared with that of TEM-1 ,-lactamasededuced from the nucleotide sequence of the ampicillin-resistance gene carried by plasmid pBR 322 (Sutcliffe,1978). TEM-1 and TEM-2 ,6-lactamases differ only by a

Vol. 251

77

M. Barth6lemy, J. Peduzzi and R. Labia -

10 20 30 40 50(A) S P Q P L E Q I K S E S Q L S G R V G M I E M D L A S G R T L T A W R A D E R F P M M S T F K V V(B) G P Q P L E Q I K Q S E S Q L S G R V G M E MDL A R T L A A W R A D E R F P M42 .S T F K V L(C) H D A E SFV L

60 70 80 90 100(A) L C G A V L A R V D AG D E Q L E RKI H YLR Q Q D VL D Y S P V S E K H L A D G M T V G E L C A A(B) L C G A V L A R V-D AG L E Q L D R R I H Y R Q Q D L V D Y S P V S E K H L V D G M TM G E L C A A(C) L C LA V L.J R V D A GR1E TLJ.KH T DGLML TVfER LL CSSLA

110 120 130 140 150(A) A I T M S D N S A A N L L L T|A|V C C P A C L T A F L R Q I C D N V T R L D R W E T E L N E A L P G(B) A I T[LS D N S A G N L L L T V G G P A C L T A F L R Q I C D N V T R L D R W E TFK L N E A L P G(C) A LP.N(C)AITMSDNATANLLLTGPETALNGHTRDWPLEILJi

160 170 180 190 200(A) D A R D T T T P A S M A A T L R K L L T S[R L S AR SQ H Q L L Q W MRV D R V AG P LI R S V L(B) D A R D T T T P A S M A A T L R K L L AC H L S A R S Q Q Q L L D W M V D D R V A G P L I RUIV L

G E L L -

210 220 230 240 250(A) P A G W F I A D K T KA G E R G A R G I V A L L G P P A S M A E R N Q(B) P A G W F I A D K T G A G E R C A R G I V A L L G P D C K P E R I V V I Y L R D T P A S M A E R N Q(C) P A G W F I A 0 A G E G I L G P 0 G K P Ft I V T T G S Q A I M V

260 265(A) Q I A G I G AJA L I E H W Q R(B) FjJI A GIC GICR(C) QI C A IS L I K ILW

Fig. 3. Amino acid sequence comparison of the three evolutionarily related fi-lactamases(A), PIT-2 f3-lactamase (present work); (B), fl-lactamase of K. pneumoniae strain LEN-I (Arakawa et al., 1986); (C), TEM-1fl-lactamase (Sutcliffe, 1978). Arakawa et al. (1986) proposed a phenylalanine residue at position 64 of the expected matureKlebsiella protein (position 85 of the pre-penicillinase), but the corresponding codon of the nucleotide sequence indicated aglutamic acid residue. Boxed residues- hold identical amino acid residues.

single amino acid substitution at position 14 (Ambler &Scott, 1978; Barthelemy et al., 1985). The PIT-2 enzymeis two residues longer at the C-terminus than is the TEM-1 protein. A 68 % amino acid sequence homology wasobserved between the two ,-lactamases. Residues 1-40,171-190 and 240-265 are poorly retained. In contrast,the sequences spanning residues 39-56, and hencesurrounding the active-site serine residue at position 45of the TEM ,B-lactamases (Fisher et al., 1980), arestrongly conserved, 17 out of 18 residues being the same.We therefore conclude that plasmid-mediated PIT-2 /1-lactamase and TEM-1 ,?-lactamase are closely evolu-tionarily related and, consequently, that the PIT-2enzyme belongs to the class A ,B-lactamases. Levesqueet al. (1987) have developed a probe to the TEM-1 ,-lactamase consisting of 424 bp from a pBR 322 DNAfragment. This probe encodes the amino acid sequencecovering residues 60-199 of the ,-lactamase. Un-expectedly, no hybridization occurred between thisprobe and the DNA of the PIT-2-,f-lactamase-encodingplasmid R974, even under low-stringency conditions.This could be indication of a greater degeneracy at theDNA level. In addition, it should be noted that Levesqueet al. (19&7) used plasmid R974 as source of PIT-2-,f-lactamase-encoding gene whereas we sequenced thep453-encoded ,B-lactamase.

Sawai et al. (1973) found some immunological homo-logy between chromosomally mediated 8-lactamase ofKlebsielta and the TEM-1 enzyme. They proposed that

the genetic determinant of the fl-lactamase was made upfrom a DNA fragment of the Klebsiella group. Thissuggestion was supported by Arakawa et al. (1986), whoprepared a 500 bp probe from the chromosomal DNAgene encoding the ,B-lactamase of K. pneumoniae strainLEN-1. The amino acid sequence predicted from thenucleotide sequence of this gene is shown in Fig. 3. Theprobe, which coded for the amino acid sequence coveringresidues 14-184, hybridized with theTEM- 1 /-lactamase-encoding plasmid pBR 322. When we compared in Fig.3 the amino acid sequence of the K. pneumoniae fi-lactamase with that of PIT-2 ,l-lactamase, we observed880 homology. No region of poorly retained sequenceappeared except at the C-terminus, the Klebsiella enzymebeing seven residues shorter than the PIT-2 protein. Thisdifference can be explained by a single-base deletionoccurring in the nucleotide sequence of the K. pneumoniaechromosome fragment prepared by Arakawa et al.(1986). Indeed, one deoxyguanine insertion at nucleotidesequence position 1117 (i.e. between the tripletsencoding Gly-256 and Gln-257 of the mature protein)would yield the following amino acid sequence:A(257) A L I E H W Q R, which fits perfectly with theC-terminus sequence of PIT-2 ,-lactamase.Our findings strongly support the hypothesis of

Nugent & Hedges (1979), who assumed that the PIT-2-,3-lactamase-encoding gene, originally carried by thechromosome in Klebsiella group, could have been laterincorporated into a plasmid. Thus the rather K.

1988

78

Amino acid sequence of PIT-2 ,-lactamase (SHV-I) 79

pneumoniae-confined PIT-2 (SHV-1) fl-lactamase couldrepresent an evolutionary step between a chromosomallyencoded /-lactamase of K. pneumoniae and the wide-spread plasmid-mediated TEM ,-lactamases.

REFERENCESAllen, G. (1981) in Sequencing of Proteins and Peptides (Work,

T. S. & Burdon, R. H., eds.), pp 223-226, Elsevier, Amster-dam, New York and Oxford

Ambler, R. P. (1965) Biochem. J. 96, 32PAmbler, R. P. (1980) Philos. Trans. R. Soc. London B 289,

321-331Ambler, R. P. & Brown, L. H. (1967) Biochem. J. 104, 784-825Ambler, R. P. & Scott, G. K. (1978) Proc. Natl. Acad. Sci.

U.S.A. 75, 3732-3736Arakawa, Y., Ohta, M., Kido, N., Fujii, Y., Komatsu, T. &

Kato, N. (1986) FEBS Lett. 207, 69-74Barthelemy, M., Peduzzi, J. & Labia, R. (1985) Ann. Inst.

Pasteur/Microbiol. 136A, 311-321Barthelemy, M., Peduzzi, J., Verchere-Beaur, C., Ben Yaghlane,

H. & Labia, R. (1986) Ann. Inst. Pasteur/Microbiol. 137B,19-27

Barthelemy, M., Pduzzi, J. & Labia, R. (1987) J. Antimicrob.Chemother. 19, 839-841

Bergstrom, S., Olsson, 0. & Nordmark, S. (1982) J. Bacteriol.150, 528-534

Fisher, J., Belasco, J. G., Khosla, S. & Knowles, J. R. (1980)Biochemistry 19, 2895-2901

Jaurin, B. & Grundstr6m, T. (1981) Proc. Natl. Acad. Sci.U.S.A. 78, 4897-4901

Kliebe, C., Nies, B. A., Meyer, J. F., Tolxdorff-Neutzling,R. M. & Wiedemann, B. (1985) Antimicrob. AgentsChemother. 28, 302-307

Levesque, R. C., Medeiros, A. A. & Jacoby, G. A. (1987) Mol.Gen. Genet. 206, 252-258

Matthew, M. (1979) J. Antimicrob. Chemother. 5, 349-358Matthew, M., Hedges, R. W. & Smith, J. T. (1979) J. Bacteriol.

138, 657-662Medeiros, A. A. (1984) Br. Med. Bull. 40, 18-27Medeiros, A. A., Cohenford, M. & Jacoby, G. A. (1985)

Antimicrob. Agents Chemother. 27, 715-719Nugent, M. E. & Hedges, R. W. (1979) Mol. Gen. Genet. 175,

239-243O'Callaghan, C. H., Simpson, I. N. & Harper, P. B. (1978)

Intersci. Conf. Antimicrob. Agents Chemother. 18th abstr.no. 91

Pitton, J. S. (1972) Rev. Physiol. Biochem. Pharmacol. 65,15-93

Sawai, T., Yamagishi, S. & Mitsuhashi, S. (1973) J. Bacteriol.115, 1045-1054

Sutcliffe, J. G. (1978) Proc. Natl. Acad. Sci. U.S.A. 75,3737-3741

Received 22 April 1987/30 September 1987; accepted 23 November 1987

Voi. 251